Implementation of Kalman filter with multicore system on chip using function — Level parallelism

Author

Majid, Mohammad Wadood ; Mirzaei, Golrokh ; Jamali, Mohsin M.

Author_Institution

Dept. of Electr. Eng. & Comput. Sci., Univ. of Toledo, Toledo, OH, USA

fYear

2013

fDate

9-11 May 2013

Firstpage

1

Lastpage

4

Abstract

Kalman filter is a very popular estimation technique used widely for linear tracking. It uses a set of noisy data as input and produces state estimates with minimum error rate. This study aims to explore how to implement implicit parallelism in multi-core processor and object tracking with task-level parallelism and Kalman Filter is parallelized on Multi-core system on chip. The novelty of this study is the introduction of Adaptive Load Balancing Approach (ALBA) to compute the nonrecursive algorithm. This approach can be applied on all form of multicore computers. The parallel Kalman Filter is developed in C# for multicore using .Net framework 4.0. It uses combination of C and CUDA for its implementation on GPU.

Keywords

C language; Kalman filters; graphics processing units; multiprocessing systems; object tracking; parallel architectures; resource allocation; .Net framework 4.0; ALBA; C; C#; CUDA; GPU; adaptive load balancing approach; compute unified device architecture; function-level parallelism; graphic processing units; multicore computers; multicore processor; multicore system-on-chip; nonrecursive algorithm; object tracking; parallel Kalman Filter; task-level parallelism; Estimation; Graphics processing units; Kalman filters; Load management; Multicore processing; Parallel processing;

fLanguage

English

Publisher

ieee

Conference_Titel

Electro/Information Technology (EIT), 2013 IEEE International Conference on

Conference_Location

Rapid City, SD

ISSN

2154-0357

Print_ISBN

978-1-4673-5207-9

Type

conf

DOI

10.1109/EIT.2013.6632705

Filename

6632705